If you measure "amount of paid work done" then most (not sure if all) experiments were failures. If you also count unpaid work, education and health benefits[1]

If you measure "amount of paid work done" then most (not sure if all) experiments were failures. If you also count unpaid work, education and health benefits[1] then things get interesting. [1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7208547/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7208547/